Background article, updated October 23, 2026. Our earlier Canadian collection has stored prices dated December 11, 2025 through March 21, 2026, with gaps. The current collection began March 22, 2026. See collection dates and gaps.
Recorded grocery prices and official inflation figures answer different questions. A pricing team may need to know what changed at a particular chain. An economist may need a measure of price change across the purchases Canadians make. Keep that difference in view before comparing the results.
Start with the question
For a product comparison, keep the product, package size, store and date together. A change in the products or stores you include can change an average even when none of the individual prices moved.
For an inflation comparison, choose the official food category and the period you want to study. A result for the stores you track should be described that way, rather than as a result for all Canadian grocery spending.
Know what the CPI measures
Statistics Canada builds the Consumer Price Index using a basket that reflects household spending. Its CPI questions and answers explain which goods and services it measures and how their weights work.
Vynn records prices from grocery listings. Those records can support a comparison of chains, products or weeks, but they do not show how much shoppers bought.
Compare the same period
Save the dates covered by each source and the date you obtained it. Check whether a file has been revised since your last analysis. When comparing results, use the same period and explain which products and stores are included.
A relationship in one period does not establish a forecasting advantage. Any forecasting result needs a separate test using only the information available at the time.
Make the work reproducible
Keep a copy of the source file, the calculation and the exclusions. Note any missing dates or changes in package size. These details let another analyst check the result.
Read how we collect prices, download the sample data, or get a free API key.